A Study of National Defense Budget Scale Forecasting Model:An Application of Time Series and Grey Prediction Model

• Main Authors: Chan, Mu-Ren, 詹沐仁
• Other Authors: Kuo, Kuo-Cheng
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/95377677285851866461

碩士 === 國防管理學院 === 國防財務資源研究所 === 95 === Defense budget has a direct impact on defense capability. A good strategic planning requires a combination of the country’s change trend and future resources allocation. A reasonable financial estimation and allocation is therefore a critical precondition for martial strategic planning. Since then, many research papers that about forecasting the scale of defense budget has based on the legal budget in the legislative. The special budget had not considered. The special budget had already been handled 36 times in Taiwan since 1950. It is handled once every 1.58 years. The special budget related to national defense affairs has been handled 8 times. It is about 22.2% that accounts for the total number of the special budget. In this study, ARIMA and GM models are adopted to forecast the scale of a defense budget. And distinguish defense budget contains the special budget or not. We compare the forecasting capabilities of two contingency both in two models. As a result, we found that defense budget does not contain special budget obtaining better prediction result. Consequently, it is unsuitable to contain the special budget while forecasting the scale of the defense budget. Under the defense budget has not contained special budget contingency, time series model is optimized for short-term forecast, as its forecasting capability drops as time increases. Grey Model only needs four sets of data, and is good and stable for short, mid, mid-and-long, and long term forecasting. At the same time, the extrapolated forecasting performance in long term by this model can be used to make up the reduced forecasting capability of traditional time series in the long term. It provides support for decision makers in long-term planning. Therefore, grey prediction model without strong hypothesis is a simple, accurate method in forecasting the scale of defense budget.